10:45 AM - 12:15 PM
[SSS04-P08] Analysis of submarine volcanic activities by using OBS waveform data
Keywords:ocean bottom seismology, seamount, seismic waveform processing, hypocenter determination
Along the Izu-Ogasawara Arc volcanic chain, there are a number of seismic events that appear to be related to volcanic activity. One of the largest events occurred on October 23, 2006 (M6.4) near Monday Seamount (Getsuyo Kaizan). In this study, we attempted to reveal detailed seismic activities (the main shock and aftershocks during the following week) near this seamount using our original method for high-precision source determination.
Until now, the event distributions has been determined only by the routine observations by the JMA and other organizations. However, they measured arrival times using the conventional manual picking even when the onset is emergent, and the station distributions are confined to the regions along the volcanic line (i.e., in the north-south direction from the epicenter), which causes poor ray coverage in the east-west direction. The resolution is therefore expected to be poor in the east-west direction, and in fact, the aftershock distribution has an east-westward spread of M8-9 class, which is inconsistent with the main shock magnitude. The analysis using new data and method is therefore necessary.
The Earthquake Research Institute had coincidentally deployed a broadband ocean bottom seismograph array in the Philippine Sea at the time of this activity (BBOBS array observations by the Stagnant Slab Project). In this study, we used data from this BBOBS array to improve coverage in the east-west direction together with advanced waveform picking techniques (Baillard et al., 2014) which are suitable for OBS data analysis. Our method has two notable features. The first is use of kurtosis to clarify the onset of OBS waveforms. The second is that we visualized the consistency among the observed travel time data and carefully excluded outliers.
Preliminary analysis revealed that the source distribution is clustered around the Monday Seamount body. The event distribution was found to form two parallel linear clusters. At the moment, only the events in the JMA catalog are analyzed, but several hundred other events have been recorded in our waveform data. We plan to analyze these events by the time of presentation and hope to reveal the detailed relationship with the magmatic activity.
Until now, the event distributions has been determined only by the routine observations by the JMA and other organizations. However, they measured arrival times using the conventional manual picking even when the onset is emergent, and the station distributions are confined to the regions along the volcanic line (i.e., in the north-south direction from the epicenter), which causes poor ray coverage in the east-west direction. The resolution is therefore expected to be poor in the east-west direction, and in fact, the aftershock distribution has an east-westward spread of M8-9 class, which is inconsistent with the main shock magnitude. The analysis using new data and method is therefore necessary.
The Earthquake Research Institute had coincidentally deployed a broadband ocean bottom seismograph array in the Philippine Sea at the time of this activity (BBOBS array observations by the Stagnant Slab Project). In this study, we used data from this BBOBS array to improve coverage in the east-west direction together with advanced waveform picking techniques (Baillard et al., 2014) which are suitable for OBS data analysis. Our method has two notable features. The first is use of kurtosis to clarify the onset of OBS waveforms. The second is that we visualized the consistency among the observed travel time data and carefully excluded outliers.
Preliminary analysis revealed that the source distribution is clustered around the Monday Seamount body. The event distribution was found to form two parallel linear clusters. At the moment, only the events in the JMA catalog are analyzed, but several hundred other events have been recorded in our waveform data. We plan to analyze these events by the time of presentation and hope to reveal the detailed relationship with the magmatic activity.